Project Summary Lung cancer is one of the most common and most deadly malignancies worldwide. Recent advances in radiation therapy treatment planning and delivery has led to the increased utilization of stereotactic body radiotherapy (SBRT) in early stage lung cancer patients. This technique delivers higher than conventional doses in only a few fractions, and has been associated with a high degree of local control. In order to ensure that the tumor is accurately treated, and to minimize the volume of healthy tissues irradiated, monitoring/tracking of the tumor position during treatment is important. Current methods rely on markers that are implanted into the patient?s tumor, or on the use of external surrogates. Both approaches have limitations. Real time x-ray imaging without implanted markers is desirable, however, the presence of bony anatomy in the thorax often obscures the view of the tumor. In this study, we propose incorporating dual energy (DE) imaging (a technique that can effectively remove overlaying bones from planar x-ray images) into the radiation therapy linac to provide high frequency, markerless motion tracking of lung tumors. The integration of DE imaging with a linear accelerator involves the following steps: A) Optimize x-ray beam quality and pulse sequence for DE imaging with template matching. A phantom will be designed that will be used to determine the optimal DE imaging parameters for template-based tracking. This phantom will also allow for quality assurance of the proposed system; B) Implement DE capability in Developer Mode on Varian linear accelerator. DE capability will be developed on a stand-alone imager/generator image, and subsequently incorporated into a linear accelerator. A workflow will be designed to allow for the clinical utilization of this system; C) Optimize performance of template-based matching using X-ray images. Methods will be identified to produce high quality templates from the treatment planning CT that can accurately track the tumor location. Additionally, phantom studies will be performed to elucidate those conditions where DE imaging is advantageous over conventional, single energy x-ray imaging; D) Evaluate improvement in tumor localization in patients using DE imaging. Improvement in the accuracy of markerless tumor tracking will be assessed in a small cohort of lung cancer patients. Since this study involves both engineering design and clinical implementation, we have assembled a group of researchers from industry (Varian Medical Systems) and academia (Loyola University Chicago), who individually have expertise in these respective areas. Through this combined effort, we expect the successful implementation of this approach will provide a practical, low cost method for enhanced tumor visualization and tracking that will improve the ability to treat lung cancer using radiation.